National Science Foundation1

BIG DATA REGIONAL INNOVATION HUBS & SPOKESUpdate on Program Activities

Fen Zhao

March 7, 2017

National Science Foundation2

KEY TAKEAWAYS

01 THE PROGRAMBrings together domain

scientists, computer scientists, and end users

to use data to solve challenges

02 THE STAKEHOLDERSEncourages collaborations with

industry, state & local governments, non profits, and others that are not typical

NSF participants

03 PARTICIPATIONOpportunity for NASA and your communities to get involved!

National Science Foundation3

in

30 mins vision of the BDHubs programactivities of funded Hubsspokes awardedopportunities for participation

National Science Foundation4

WHAT IS THE HISTORY BEHIND BDHUBS?The National Big Data R&D Initiative & Data to Knowledge to Action (Data2Action)

MAR2012

LaunchNITRD Agencies (lead by NSF) kick off the National Big Data R&D Initiative with new federal programs totaling $200M

MAY2013

Big Data Partnerships WorkshopIndustry, academia, and government representatives gathered to learn about current Big Data partnership and brainstorm new ideas

NOV2013

Data2Action90 organizations announce 29 new Big Data partnerships supported by $100M in non-federal funds

JUN2014

Partnerships Bear FruitPartnerships update NITRD on midterm outcomes from announced projects

MAR2015

BDHubsNSF initiates BDHubseffort to sustain and scale up collaborative Big Data innovation activities

National Science Foundation5

THE HISTORY BEHIND BDSPOKESBD Spokes is the second phase of a long term NSF agenda for Big Data Partnerships

MAR2015

BD Hubs LaunchedBD Hubs solicitation to fund four regional Hubs is released

APR2015

Big Data Regional Charrettes HeldIndustry, academia, and government representatives gathered in four charrettes around the country

SEPT2015

Hubs Awards MadeAwards made to coordinating institutions

NOV2015

BD SpokesBD Spokes solicitation released before 5th

DC national charrette (bdhubs.info)

SEPT2016

BD Spokes Awarded10 (+1) Spokes and 10 planning grants awarded

National Science Foundation6

WHAT IS THE BDHUBS NETWORK?“Hub and Spoke”– A Nation-Wide Network for Data Innovation

1 HubsLocal stakeholders

guide activities locally and nationally

2Spokes

Hub selects somelocal priority areas(i.e. transportation,

manufacturing)

3 NodesPartnerships formed

to drive specific end goals in priority areas

National Science Foundation7

WITHIN THE BIG DATA PORTFOLIO OF PROGRAMS

Within the broader portfolio, BD Hubs and BD Spokesfocuses on building partnerships around Big Data

RESEARCHCritical Techniques & Technologies for … Big Data (BIGDATA)

INFRASTRUCTUREData Infrastructure Building Blocks (DIBBS)

EDUCATIONNational Research Traineeship (NRT)

PARTNERSHIPSBig Data Regional Innovation Hubs: Spokes (BD Spokes)

National Science Foundation8

BD HubsFounding organizations for BDHubs in 2015Points indicate affiliations of individuals named as steering council members and/or task leads or senior personnel.

University

HPC Center

Non-profit

Government

Industry

MIDWEST106 Personnel79 Organizations12 states

UND(co-PI)

Iowa State (co-PI)

UIUC/NCSA (PI)Indiana U (co-PI)

U of M (co-PI)

NORTHEAST193 Personnel99 Institutions9 States

Columbia (PI)

WEST86 Personnel 47 Organizations13 States

UW (PI)

Berkeley (PI)

UCSD/SDSC (PI)

SOUTH116 Personnel95 Organizations15 States + DC

UNC/RENCI (PI)

Georgia Tech (PI)

Alaska & Hawaii are part of the West regionUS Territories can participate in any region

National Science Foundation9

HUB ACTIVITIESHubs ideate and coordinate Spokes, but also host a variety of activities for the community

Microsoft awards Hubs $3M in cloud computing credits

Massive regional All-Hands with

hundred of attendees

Early career researcher programs with CCC 3 years

sociotechnical study of Hubs

National Science Foundation1010

The strategy behind

BD SPOKES

BD Spokes are not your typical R&D project

nor are they mini Hubs

National Science Foundation11

MISSION DRIVEN SPOKESBD Spokes proposals must articulate a clear focus within a specific Big Data topic or application area, while highlighting their Big Data Innovation theme.

All BD Spokes must have clearly defined mission statements with goals and corresponding metrics of success.

National Science Foundation12

SPOKESMAJORTHEMESThree different ways of slicing the Big Data Innovation problem

SPOKES TO DIRECTLY ADDRESS

National Science Foundation13

AREAS OF EMPHASISSome NSF priority areas include

NEUROSCIENCE REPLICABILITY & REPRODUCABILITYIN DATA SCIENCE

SMART & CONNECTED COMMUNITIES

DATA PRIVACY DATA INTENSIVE RESEARCH IN THE SOCIAL, BEHAVIORAL, & ECONOMIC SCIENCES

EDUCATION

National Science Foundation14

Percent funding per region

West18%

South26% North

east28%

Mid west28%

Percent funding per topic area

Cybersecurity2%

Material Science8%

Neuroscience8%

Education9%

Environment17%

Sharing and Reproducibility18%

Health18%

Smart Cities20%

Total Spokes ~$12M in first round

National Science Foundation15

BD Spokes:Phase 1Includes lead and non-lead institutions for Spokes and Planning Grants

Planning Grant LeadPlanning Grant Non-leadSpoke Lead

Spoke Non-Lead or Subaward

MIDWEST

NORTHEAST

WEST

SOUTH

Alaska & Hawaii are part of the West regionUS Territories can participate in any region

16

IBM WATSON + ENCYCLOPEDIA OF LIFE“Using Big Data for Environmental Sustainability: Big Data + AI Technology = Accessible, Usable, Useful Knowledge!”

Encyclopedia of Life (EOL) is the world's largest database of biological species and other biodiversity information. EOL also works closely with scores of other biodiversity datasets such as BISON, GBIF, and OBIS.

This project seeks to make EOL and related biodiversity data sources accessible, usable, and useful, by integrating extant artificial intelligence tools for information extraction, modeling and simulation, and question answering.

(1) Cognopsi: semantically annotate documents in EOL through controlled vocabularies for specific domains within ecological and environmental science

(2) MILA-S: constructs conceptual models of ecological phenomena and automatically spawns simulation models; use with EOL TraitBank, to generate and test explanatory hypotheses as well as make predictions about ecosystems

(3) Watson+: adds semantic processing to Watson to act as a virtual research assistant; will train Watson+ for answering questions about biological species using EOL.

Georgia Tech & Smithsonian InstitutionLead Proposal: 1636848

17

SMART GRID DATA SHARING“Smart Grids Big Data”

Will create an organization that brings together a cross disciplinary capability from academia, industry, and government. The goal of the project is to ideate from Smart Grid Data new knowledge and solutions offering major improvements in smart grid operation (e.g., power generation and distribution; renewable energy) and smart grid user necessities (critical infrastructures, smart cities, transportation, etc.)

Over 67 organizations submitted letters of collaboration.

Will be building an open data and software exchange. Initial data committed:

• data provided by over 50 utility companies and 30 utility industry solution vendors

• National Lightning Detection Network Data from Vaisala

• Lawrence Livermore National Lab (LLNL) data coming from local sensor network including several PMU’s and weather monitoring devices

• International partners: Brazilian power system project MedFasee; demand side management studies University of Manchester, renewable generation data collection activities -University of Cyprus

• And many, many more

Texas A&M et al.Lead Proposal:1636772

18

DIGITALAGRICULTURE“Unmanned Aircraft Systems (UAS), Plant Sciences and Education”

Will organize academic, industrial, and governmental sectors around the development of policies and best practices for data science and Big Data applications in agriculture

Main focus on automating the Big Data lifecycle:

• automation of transport, storage, dissemination, and analysis of UAS imagery and ground characterizations

• automation of Big Data pipelines and the integration, interoperability and re-use of databases across plant and cropping systems – from farm management and remote sensing to high throughput plant phenomics and crop genomics

Activities focus on workshop series, hackathons, challenges, for example:

• Will develop a set of webinars on ontology, analytics, data management, data sharing, data standards and conventions, and data instrumentation to be used as a blueprint for a graduate level seminar on data science in agriculture

• Runs a competition for “mini proposals” in data annotation and interoperability for ag-genomics

University of North DakotaProposal: 1636865

National Science Foundation19

KEY TAKEAWAYS

01 THE PROGRAMBrings together domain

scientists, computer scientists, and end users

to use data to solve challenges

02 THE STAKEHOLDERSEncourages collaborations with

industry, state & local governments, non profits, and others that are not typical

NSF participants

03 PARTICIPATIONOpportunity for NASA and your communities to get involved!

National Science Foundation20

FOR FURTHER QUESTIONS CONTACTFen Zhao, fzhao@nsf.gov 703 292 7344

NSF Headquarters, Arlington VA